AMD Brings Smart Access Memory (Resizable Bar) To Ryzen 3000 Desktop CPUs, Up To 16% Performance Boost In AAA Games

Submit

In addition to the Radeon RX 6700 XT unveil, AMD also announced that is bringing Smart Access Memory (Resizable Bar) technology to its Ryzen 3000 Desktop CPUs based on Zen 2 cores. This is great news for a large majority of PC users who are running Matisse CPUs and the extra boost in performance thanks to BAR support is much welcomed by the gaming community.

AMD Smart Access Memory (Resizable Bar) Now Available on Ryzen 3000 'Matisse Zen 2' Desktop CPUs

AMD Smart Access Memory or Resizable BAR technology is great when it works and we have seen now in multiple performance tests that the gains are there. It's definitely not across the board but in certain titles, you can get a big boost in performance for free. For Ryzen 3000, AMD has stated that users can expect a performance boost of up to 16% which is pretty impressive. AMD showed the performance improvement with its own Radeon RX 6000 series class GPUs but we know that motherboard manufacturers have opened BAR support for NVIDIA GPUs on AMD motherboards too.

ASUS’s ProArt StudioBook Pro 16 With AMD Ryzen 9 5900HX & NVIDIA RTX 3070 Spotted

The AMD Ryzen 3000 series family launched alongside the 400-series platform and those already have BIOS support for resizable-BAR. However, we can expect a more formal BIOS rollout that will fully enable Ryzen 3000 CPU support on both AMD 500 & 400 series motherboards. Intel platforms also opened BAR support down to 300-series motherboards which goes all the way back to 8th Gen CPUs so it was wise for AMD to make this move.

Just to recap, AMD's Smart Access Memory (Resizable Bar) technology does not require you to invest in a PCIe Gen 4 platform as it will be supported by PCIe Gen 3 too. Based on what we know so far, BAR essentially defines how much discrete GPU memory space can be mapped and today's PCs are typically limited to 256 MB of mapped memory. AMD claims that with SAM, they can access all of the GPU memory, removing any bottlenecks to allow for faster performance.

Submit